Google Gemini can now be integrated into Android apps

Google steps on the gas pedal Gemini, its latest generation of artificial intelligence. Although this technology was he announced As recently as last week, the company does not want to waste time and its intention is that more and more users use it. To support their expansion, they rely on one of their most used products: Android.

Google just announced that Starting today, developers can integrate Gemini into their applications. Businesses can also use it in their workflows. That’s what they have released Gemini API, a set of tools now available through Google AI Studio and Google Cloud Vertex AI.

It is important to note that the API currently only offers access to solutions Gemini Pro. That is, the proposal, the processing of which takes place on Google’s servers, not on the user’s device.

Developers can thus create applications that offer e.g. text generation function. We must remember that Gemini is a multimodal language model. This means that it can generate content based on a user request, but also interpret existing content – both text and images.

Google Gemini |  Artificial Intelligence

Since the processing takes place in Google data centers, Gemini Pro integration is not always free. To avoid logging out, the API must not exceed 60 queries per minute.

If the app needs more, the developers will have to pay based on the requests raised. The pricesnext:

Data input:

$0.00025 for every 1000 characters entered.
$0.0025 for each image entered.

Data output:

$0.0005 per 1000 characters of output.

“In the version available today, Gemini Pro accepts text as input and generates text as output. We’ve also made available the multimodal Gemini Pro Vision terminal, which accepts text and images as input and text as output.”


An important point is that it is not possible to use a payment proposal for now. Developers are therefore limited to using the free plan. It is certainly a temporary measure until the community learns to use the API and performs the first tests. 60 queries per minute is more than enough for experimentation.

As of this writing, Gemini Pro has support for 38 languages. As a result, it can meet the needs of developers – and users – in more than 180 countries.

Google is betting everything on Gemini

No doubt about it Gemini became a priority for Google. Advances by its competitors, notably OpenIA — which works closely with Microsoft — have caused Mountain View to redouble its AI efforts.

With the launch of the API for integrating Gemini Pro into third-party applications, a world of possibilities opens up, the limit of which will be the creativity of the developers themselves. And be careful because this is just Google’s first step. After Android, Google Chrome and Firebase will be the next to adopt Gemini.

Google also plans to launch next year Gemini Ultra, a larger model capable of performing more complex tasks. Of course, they must first ensure that the existing designs, both Gemini Nano and Gemini Pro, meet the demands of users, companies and developers. In fact, they mention that community comments will be key to further improving their technology.

Receive from us every morning reporter. A guide to understanding what matters in relation to technology, science and digital culture.


Ready! You have already subscribed to this subscription

An error occurred, please refresh the page and try again

Also in Hipertextual: